R
Roberto Calandra
Researcher at Facebook
Publications - 84
Citations - 4252
Roberto Calandra is an academic researcher from Facebook. The author has contributed to research in topics: Reinforcement learning & Tactile sensor. The author has an hindex of 24, co-authored 79 publications receiving 2448 citations. Previous affiliations of Roberto Calandra include Technische Universität Darmstadt & University of California, Berkeley.
Papers
More filters
Proceedings Article
Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models
TL;DR: This paper proposes a new algorithm called probabilistic ensembles with trajectory sampling (PETS) that combines uncertainty-aware deep network dynamics models with sampling-based uncertainty propagation, which matches the asymptotic performance of model-free algorithms on several challenging benchmark tasks, while requiring significantly fewer samples.
Posted Content
Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models
TL;DR: This article proposed a probabilistic ensembles with trajectory sampling (PETS) algorithm, which combines uncertainty-aware deep network dynamics models with sampling-based uncertainty propagation to match the asymptotic performance of model-based and model-free deep RL algorithms.
Journal ArticleDOI
Bayesian optimization for learning gaits under uncertainty
TL;DR: Bayesian optimization, a model-based approach to black-box optimization under uncertainty, is evaluated on both simulated problems and real robots, demonstrating that Bayesian optimization is particularly suited for robotic applications, where it is crucial to find a good set of gait parameters in a small number of experiments.
Posted Content
Learning Invariant Representations for Reinforcement Learning without Reconstruction
TL;DR: This work studies how representation learning can accelerate reinforcement learning from rich observations, such as images, without relying either on domain knowledge or pixel-reconstruction, and proposes a method to learn robust latent representations which encode only the task-relevant information from observations.
Journal ArticleDOI
DIGIT: A Novel Design for a Low-Cost Compact High-Resolution Tactile Sensor With Application to In-Hand Manipulation
Mike Lambeta,Po-Wei Chou,Stephen Tian,Brian Yang,Benjamin Maloon,Victoria Rose Most,Dave Stroud,Raymond Santos,Ahmad Byagowi,Gregg Kammerer,Dinesh Jayaraman,Roberto Calandra +11 more
TL;DR: The DIGIT sensor is introduced, an inexpensive, compact, and high-resolution tactile sensor geared towards in-hand manipulation that is demonstrated by training deep neural network model-based controllers to manipulate glass marbles in- hand with a multi-finger robotic hand.